A parametric environmental impact model for manufacturing components based on machine learning techniques

Environmental sustainability-oriented design is becoming increasingly important in the industrial field partly because of the effects of climate change. Sustainable development-oriented choices are most effective at the early design stage. The design team must be able to assess approximately and quickly the environmental impact early in the design phase. From these motivations comes the need for a method that quickly and with few parameters can estimate the product environmental impact during the conceptual design phase. Machine learning techniques appear to be well suited to meet this challenge. Machine learning is an established research topic in Industry 4.0 and its adoption is increasing. The integration of machine learning within conceptual design quickly facilitates the approximate assessment of environmental impact through high-level data. In this paper, a method is proposed to obtain a parametric model for the environmental impact assessment of manufacturing components at the early design stage. It allows consistent considerations concerning environmental matters, albeit little information available during design phase.